CN108416155A - A kind of microwave gallium nitride device physical base large-signal model method for building up and system - Google Patents
A kind of microwave gallium nitride device physical base large-signal model method for building up and system Download PDFInfo
- Publication number
- CN108416155A CN108416155A CN201810227481.8A CN201810227481A CN108416155A CN 108416155 A CN108416155 A CN 108416155A CN 201810227481 A CN201810227481 A CN 201810227481A CN 108416155 A CN108416155 A CN 108416155A
- Authority
- CN
- China
- Prior art keywords
- signal model
- physical
- parameter
- base large
- physical base
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/36—Circuit design at the analogue level
- G06F30/367—Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/36—Circuit design at the analogue level
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/18—Manufacturability analysis or optimisation for manufacturability
Landscapes
- Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Crystals, And After-Treatments Of Crystals (AREA)
- Junction Field-Effect Transistors (AREA)
Abstract
The present invention discloses a kind of microwave gallium nitride device physical base large-signal model method for building up and system.This method includes：It obtains to test result data of each gallium nitride device under different input power in same processing line；Calculate the mean value of the test result data under each input power；Physical base large-signal model is established based on the mean value, and it is nominal value that physical parameter, which is arranged, and then obtains fitting parameter initial value；Screen top edge curve and the corresponding device of lower edge curve；Establish the physical base large-signal model of top edge device and lower edge device；Fitting parameter is substituted into the physical base large-signal model of two edge devices, fitting parameter and physical parameter are adjusted, ensure that two models share same group of fitting parameter and physical parameter is different, so that it is determined that the fitting parameter and physical parameter of physical base large-signal model.The solution of the present invention can improve the accuracy of the confidence level and physical parameter of fitting parameter extraction, to improve the precision of model.
Description
Technical field
The present invention relates to GaN high electron mobility transistor technical fields, more particularly to a kind of microwave gallium nitride device
Part physical base large-signal model method for building up and system.
Background technology
GaN high electron mobility transistor (GaN HEMT) is excellent due to its high frequency, high power density, high pressure resistant etc.
Characteristic has extremely important application in microwave and millimeter wave solid state power circuit.In order to establish technological parameter and device microwave
The contact of characteristic carries out technology stability and circuit yield analysis, and establishing accurate large-signal model for GaN HEMT is
It is highly desirable.The existing excessively many relevant reports of experience base large-signal model of GaN HEMT, experience base large-signal model
Modeling technique has also reached its maturity.But most of experience base large-signal models include a large amount of fitting parameter, model parameter it is more
Value sex chromosome mosaicism can make model parameter lose physical significance, often be brought greatly to more biasing applicabilities of model, broadband character
Challenge, to increase the difficulty of model parameter extraction；In addition, experience base large-signal model parameter usually lacks physical significance,
It is difficult to intuitively be contacted with device foundation, to cannot achieve to device technology and make offer guidance.Different from traditional warp
Base large-signal model is tested, physical base large-signal model is derived by by the physical parameter of device, therefore model parameter has mostly
There is physical significance, directive function can be played to device technology and processing and fabricating, while the convergence problem of model also obtains
Larger improvement.Therefore, nearly 2 years, physical base large-signal model was just gradually replacing traditional experience base large-signal model, becomes
The research hotspot of non-linear large signal modeling.
Although physical base large-signal model fitting parameter is less, the reasonability of fitting parameter will significantly influence
The extraction of model physical parameter, to which accurate description, therefore physical base large-signal model can not be carried out to the truth of device
The accurate extraction of fitting parameter is just particularly important.Although having there are many papers to carry out model parameter extraction at present to grind
Study carefully, but be mostly experience base large-signal model parameter extracting method introduction, for GaN HEMT device physical base large-signal models
Parameter especially fitting parameter extracting method it is actually rare.
The physical base large-signal model based on surface potential that Wu Qingzhi of University of Electronic Science and Technology et al. was proposed in 2016 leads to
It crosses solution surface potential and obtains intrinsic region channel charge and drain current model.The model really reflects the working condition of device,
It is preliminary to establish contacting for physical parameter and device property, but and unresolved fitting parameter accurate extraction problem.
The SheikhAamirAhsan in your branch school of the Institute of Technology of India Kemp et al. is before being based on them in 2017 in table
Work in terms of the gesture physical base modeling of face, it is proposed that a set of more biasing large-signal model parameter extraction flows are different from traditional
Experience base and artificial neural network basic mode type, the parameter extraction flow do not need complicated optimization algorithm, only by being fitted device
DC-IV characteristics the large-signal model parameter extraction under more bias conditions can be completed.Although the work is directed to surface potential physics
Basic mode type, but and unresolved fitting parameter accurate extraction problem.
Invention content
The object of the present invention is to provide a kind of microwave gallium nitride device physical base large-signal model method for building up and systems, carry
The accuracy of high fitting parameter extraction, to improve the accuracy of model.
To achieve the above object, the present invention provides following schemes：
A kind of microwave gallium nitride device physical base large-signal model method for building up, including：
Each microwave gallium nitride device to be measured in same processing line is tested in acquisition under different input power
Obtained test result data；
Calculate the mean value of the test result data under each input power；
Mean value based on the test result data establishes initial physical base large-signal model, and the initial physical is arranged
The physical parameter of base large-signal model is nominal value, obtains fitting parameter initial value；
The test result data for screening each microwave gallium nitride device is formed by curve and envelope top edge song
The hithermost curve of line and with the hithermost curve of envelope lower edge curve, obtain top edge device curves and lower edge device be bent
Line；The top edge device is the microwave gallium nitride device corresponding to the top edge device curves；The lower edge device
Part is the microwave gallium nitride device corresponding to the lower edge device curves；
Establish the first physical base large-signal model based on the top edge device and the based on the lower edge device
Two physical base large-signal models；
The fitting parameter initial value is substituted into the first physical base large-signal model and the big signal of the second physical base
Model, and fitting parameter to the first physical base large-signal model and physical parameter and the big signal of the second physical base
The fitting parameter and physical parameter of model are finely adjusted, and ensure the first physical base large-signal model and second physics
Base large-signal model shares same group of fitting parameter and physical parameter is different, makes simulation result and the top edge device and described
Deviation between the test result data of lower edge device meets predetermined threshold value, to obtain target fitting parameter and target physical
Parameter.
Optionally, the fitting parameter initial value is substituted into the first physical base large-signal model and described second described
Physical base large-signal model, and the fitting parameter to the first physical base large-signal model and physical parameter and described second
The fitting parameter and physical parameter of physical base large-signal model are finely adjusted, and ensure the first physical base large-signal model and
The second physical base large-signal model shares same group of fitting parameter and physical parameter is different, makes simulation result and the top
Deviation between edge device and the test result data of the lower edge device meets predetermined threshold value, to obtain target fitting ginseng
Number and target physical parameter further include：
Two are arbitrarily chosen from the gallium nitride device in addition to the top edge device and the lower edge device
Gallium nitride device obtains two gallium nitride devices to be verified；
The physical base large-signal model for establishing two gallium nitride devices to be verified obtains the big signal mode of third physical base
Type and the 4th physical base large-signal model；
The target fitting parameter is substituted into the third physical base large-signal model respectively and the 4th physical base is big
Signal model, the physics by only adjusting the third physical base large-signal model and the 4th physical base large-signal model are joined
Number, obtains the simulation result of the third physical base large-signal model and the 4th physical base large-signal model；
Judge the simulation result difference of the third physical base large-signal model and the 4th physical base large-signal model
Whether the deviation between corresponding test result data is satisfied by predetermined threshold value；
If at least there is an object in the third physical base large-signal model and the 4th physical base large-signal model
The simulation result for managing base large-signal model still cannot be satisfied the predetermined threshold value after only to reason parameter adjustment, then adjusts again
The whole target fitting parameter；
If the simulation result of the third physical base large-signal model and the 4th physical base large-signal model is passing through
It is satisfied by the predetermined threshold value after physical parameter adjustment, then is established with batch gallium nitride device using the target fitting parameter
Physical base model, to instruct device design.
Optionally, the fitting parameter initial value is substituted into the first physical base large-signal model and second physical base
Large-signal model, and fitting parameter to the first physical base large-signal model and physical parameter and second physical base
The fitting parameter and physical parameter of large-signal model are finely adjusted, and ensure the first physical base large-signal model and described
Two physical base large-signal models share same group of fitting parameter and physical parameter is different, make simulation result and the top edge device
Deviation between the test result data of the lower edge device meets predetermined threshold value, to obtain target fitting parameter and mesh
Physical parameter is marked, is specifically included：
The fitting parameter initial value is substituted into the first physical base large-signal model and the second physical base large-signal model
In；
The physical parameter of the top edge device and the lower edge device is adjusted in default adjusting range, described in judgement
The simulation result of first physical base large-signal model and the second physical base large-signal model respectively with corresponding test result
Whether the deviation between data is satisfied by predetermined threshold value, obtains judging result；
If the judging result indicates that the deviation meets the predetermined threshold value, terminate parameter adjustment, determines described quasi-
It is target fitting parameter and target physical parameter to close the physical parameter after initial parameter values and fine tuning；
If the judging result indicates that the deviation exceeds the predetermined threshold value, the fitting parameter initial value is finely tuned, is made
The simulation result of the first physical base large-signal model and the second physical base large-signal model respectively with corresponding test
Deviation between result data is satisfied by predetermined threshold value, and determines that the fitting parameter after fine tuning and the physical parameter are mesh
Mark fitting parameter and target physical parameter.
Invention additionally discloses a kind of microwave gallium nitride device physical base large-signal models to establish system, including：
Acquisition module, for obtaining to each microwave gallium nitride device to be measured in same processing line in different input works
The test result data tested under rate；
Mean value computation module, the mean value for calculating the test result data under each input power；
Fitting module is established initial physical base large-signal model for the mean value based on the test result data, and is set
The physical parameter for setting the initial physical base large-signal model is nominal value, obtains fitting parameter initial value；
Edge devices screening module, the test result data for screening each microwave gallium nitride device are formed by
In curve with the hithermost curve of envelope top edge curve and with the hithermost curve of envelope lower edge curve, obtain top edge device
Part curve and lower edge device curves；The top edge device is the microwave nitridation corresponding to the top edge device curves
Gallium device；The lower edge device is the microwave gallium nitride device corresponding to the lower edge device curves；
First model building module, for establishing the first physical base large-signal model and base based on the top edge device
In the second physical base large-signal model of the lower edge device；
Model parameter determining module, for by the fitting parameter initial value substitute into the first physical base large-signal model and
The second physical base large-signal model, and fitting parameter to the first physical base large-signal model and physical parameter and
The fitting parameter and physical parameter of the second physical base large-signal model are finely adjusted, and ensure the big letter of the first physical base
Number model and the second physical base large-signal model share same group of fitting parameter and physical parameter is different, make simulation result with
Deviation between the top edge device and the test result data of the lower edge device meets predetermined threshold value, to obtain mesh
Mark fitting parameter and target physical parameter.
Optionally, the system also includes：
Random screening module is used for from the gallium nitride device in addition to the top edge device and the lower edge device
Two gallium nitride devices are arbitrarily chosen in part, obtain two gallium nitride devices to be verified；
Second model building module, the physical base large-signal model for establishing two gallium nitride devices to be verified,
Obtain third physical base large-signal model and the 4th physical base large-signal model；
Module is substituted into, for the target fitting parameter to be substituted into the third physical base large-signal model and described respectively
4th physical base large-signal model, by only adjusting the third physical base large-signal model and the big signal of the 4th physical base
The physical parameter of model obtains the emulation knot of the third physical base large-signal model and the 4th physical base large-signal model
Fruit；
Judgment module, for judging the third physical base large-signal model and the 4th physical base large-signal model
Whether the deviation between corresponding test result data is satisfied by predetermined threshold value to simulation result respectively；
Feedback module, if being used in the third physical base large-signal model and the 4th physical base large-signal model extremely
It still cannot be satisfied after only to physical parameter adjustment in the presence of the simulation result of a physical base large-signal model less described pre-
If threshold value, then the target fitting parameter is readjusted；
Application module, if for the imitative of the third physical base large-signal model and the 4th physical base large-signal model
True result is being satisfied by the predetermined threshold value after passing through physical parameter adjustment, then establishes same batch using the target fitting parameter
The physical base model of secondary gallium nitride device, to which instruct device designs.
Optionally, the model parameter determining module, specifically includes：
Parameter substitutes into unit, for the fitting parameter initial value to be substituted into the first physical base large-signal model and described second
In physical base large-signal model；
Physical parameter adjustment unit, for adjusting the top edge device and the lower edge device in default adjusting range
The physical parameter of part judges the simulation result of the first physical base large-signal model and the second physical base large-signal model
Whether the deviation between corresponding test result data is satisfied by predetermined threshold value respectively, obtains judging result；
First parameter determination unit is tied if indicating that the deviation meets the predetermined threshold value for the judging result
Beam parameter adjustment determines that the physical parameter after the fitting parameter initial value and fine tuning is target fitting parameter and target physical
Parameter；
Second parameter determination unit, it is micro- if indicating that the deviation exceeds the predetermined threshold value for the judging result
The fitting parameter initial value is adjusted, the emulation of the first physical base large-signal model and the second physical base large-signal model is made
As a result the deviation respectively between corresponding test result data is satisfied by predetermined threshold value, and determines the fitting ginseng after fine tuning
Number and the physical parameter are target fitting parameter and target physical parameter.
According to specific embodiment provided by the invention, the invention discloses following technique effects：Method disclosed by the invention
And system, using the mean value of test result data, fixed physical parameter extraction fitting parameter had both weakened the physics of fitting parameter
Meaning, and the higher fitting parameter initial value of confidence level can be obtained, to which fitting parameter extraction can be improved in the elementary step
Reasonability.Meanwhile the gallium nitride device that the curve for choosing test result data is located at envelope edge both ensured as modeled device
The fitting parameter of extraction is unrelated with device technology, and can extract accurate physical parameter, improves parameter extraction
The accuracy of efficiency and model foundation.
Description of the drawings
It in order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, below will be to institute in embodiment
Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations of the present invention
Example, for those of ordinary skill in the art, without having to pay creative labor, can also be according to these attached drawings
Obtain other attached drawings.
Fig. 1 is the method flow diagram of microwave gallium nitride device physical base large-signal model method for building up embodiment of the present invention；
Fig. 2 is the fitting parameter initial value of microwave gallium nitride device physical base large-signal model method for building up embodiment of the present invention
The simulation result and device measured data mean value contrast curve of corresponding large-signal model；
Fig. 3 is the top edge of the screening of microwave gallium nitride device physical base large-signal model method for building up embodiment of the present invention
The curve graph of device curves and lower edge device curves；
Fig. 4 is the edge devices physics of microwave gallium nitride device physical base large-signal model method for building up embodiment of the present invention
The simulation result of base large-signal model and measured result contrast curve；
Fig. 5 is two arbitrarily chosen of microwave gallium nitride device physical base large-signal model method for building up embodiment of the present invention
The simulation result of a gallium nitride device and measured result comparison diagram；
Fig. 6 is the system construction drawing that microwave gallium nitride device physical base large-signal model of the present invention establishes system embodiment.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
The present invention basic thought be：The test result data of the gallium nitride device in same processing line is obtained first；
Secondly, test data mean value, fixed physical parameter extraction fitting parameter are calculated；Then, envelope edge devices, fixed fitting are chosen
Parameter extraction physical parameter；Finally, the reasonability that fitting parameter is verified with gallium nitride device, to verify model.
In order to make the foregoing objectives, features and advantages of the present invention clearer and more comprehensible, below in conjunction with the accompanying drawings and specific real
Applying mode, the present invention is described in further detail.
Fig. 1 is the method flow diagram of gallium nitride device physical base large-signal model method for building up embodiment of the present invention.
Referring to Fig. 1, the gallium nitride device physical base large-signal model method for building up, including：
Step 101：It obtains to each microwave gallium nitride device to be measured in same processing line in different input power
The lower test result data tested.Mainly for the gallium nitride technique of 0.25 μm of grid length in the embodiment of the present invention
The GaN high electron mobility transistor (GaN HEMT) of line production is tested.The optional DC test of test method and/or
Radio frequency testing.
Step 102：Calculate the mean value of the test result data under each input power.Because of sample device
(gallium nitride device for measurement chosen) randomly selects on same gallium nitride line of production technology.So sample
Device corresponds to the physical parameter of physics basic mode type also with processing line random fluctuation.The test result of sample device can thus be used
The mean value of data is as research object.
Step 103：Mean value based on the test result data extracts fitting parameter.When with the mean value of test result data
When determining fitting parameter when as research object, the physical parameter for the physical base large-signal model established according to the fitting parameter can
To be set as nominal value.Physical base large-signal model can be established on this basis, obtain rational fitting parameter initial value.
Fig. 2 is that the fitting parameter initial value of gallium nitride device physical base large-signal model method for building up embodiment of the present invention corresponds to
The simulation result of large-signal model and device measured data mean value contrast curve.
Referring to Fig. 2, from figure 2 it can be seen that output power, the work(of the obtained fitting parameter initial value of method of the present invention
Rate gain and power added efficiency are fitted preferably with input power.Illustrate the confidence level for the fitting parameter initial value that the present invention obtains
It is higher.
Step 104：The test result data for screening each microwave gallium nitride device is formed by curve and envelope
The hithermost curve of top edge curve and with the hithermost curve of envelope lower edge curve, obtain top edge device curves and following
Edge device curves；The top edge device is the microwave gallium nitride device corresponding to the top edge device curves；It is described
Lower edge device is the microwave gallium nitride device corresponding to the lower edge device curves.
Above-mentioned steps 104 are to extract the process of edge devices and edge devices curve.Extract edge devices and edge devices
Curve is to determine that physical parameter lays the foundation.After the present invention obtains fitting parameter initial value in step 103, physical parameter is just
It is easier to be extracted.And physical parameter exactly monitors processing line, the key factor of optimised devices.In order to which accurate extracts physical is joined
Number, the embodiment of the invention choose two devices of envelope edge as research object from test result data.
Fig. 3 is the top edge device of the screening of gallium nitride device physical base large-signal model method for building up embodiment of the present invention
The curve graph of curve and lower edge device curves.
Referring to Fig. 3, top edge device is T in figure, and lower edge device is B.It extracts according to this embodiment of the invention
The top edge device T and lower edge device B arrived has higher independence and representativeness.
Step 105：It establishes the first physical base large-signal model based on the top edge device and is based on the lower edge
Second physical base large-signal model of device.Keysight-ADS simulation softwares are specially utilized, in the same circuit diagram
In, the physical base large-signal model of the top edge device and the lower edge device is established, can ensure two models in this way
Share same group of fitting parameter and physical parameter is different.
Step 106：The fitting parameter initial value is substituted into the first physical base large-signal model and second physics
Base large-signal model, and fitting parameter to the first physical base large-signal model and physical parameter and second physics
The fitting parameter and physical parameter of base large-signal model are finely adjusted, and ensure the first physical base large-signal model and described
Second physical base large-signal model shares same group of fitting parameter and physical parameter is different, makes simulation result and the top edge device
Deviation between part and the test result data of the lower edge device meets predetermined threshold value, to obtain target fitting parameter and
Target physical parameter.The step 106 specifically includes：
The fitting parameter initial value is substituted into the first physical base large-signal model and the second physical base large-signal model
In.This method of the present invention had both been convenient for the physical base large-signal model of two edge devices of setting to share same group of fitting parameter,
It can make physical parameter different with device again and fluctuate.
The physical parameter of the top edge device and the lower edge device is adjusted in default adjusting range, described in judgement
The simulation result of first physical base large-signal model and the second physical base large-signal model respectively with corresponding test result
Whether the deviation between data is satisfied by predetermined threshold value, obtains judging result.
If the judging result indicates that the deviation meets the predetermined threshold value, terminate parameter adjustment, determines described quasi-
It is target fitting parameter and target physical parameter to close the physical parameter after initial parameter values and fine tuning；
If the judging result indicates that the deviation exceeds the predetermined threshold value, the fitting parameter initial value is finely tuned, is made
The simulation result of the first physical base large-signal model and the second physical base large-signal model respectively with corresponding test
Deviation between result data is satisfied by predetermined threshold value, and determines that the fitting parameter after fine tuning and the physical parameter are mesh
Mark fitting parameter and target physical parameter.
Fig. 4 is that the edge devices physical base of gallium nitride device physical base large-signal model method for building up embodiment of the present invention is big
The simulation result of signal model and measured result contrast curve.
Referring to Fig. 4, the curve in figure is simulation result, and square is the test result data of top edge device, triangle
For the test result data of lower edge device.Figure 4, it is seen that the obtained fitting parameter of the embodiment of the present invention
Accuracy is higher, and the accuracy of physical base large-signal model is higher.
Step 107：It is arbitrary from the gallium nitride device in addition to the top edge device and the lower edge device
Two gallium nitride devices are chosen, two gallium nitride devices to be verified are obtained；
Step 108：The physical base large-signal model for establishing two gallium nitride devices to be verified, obtains third physical base
Large-signal model and the 4th physical base large-signal model；
Step 109：The target fitting parameter is substituted into the third physical base large-signal model and the described 4th respectively
Physical base large-signal model only adjusts the object of the third physical base large-signal model and the 4th physical base large-signal model
Parameter is managed, the simulation result of the third physical base large-signal model and the 4th physical base large-signal model is obtained；
Step 110：Judge the emulation of the third physical base large-signal model and the 4th physical base large-signal model
As a result whether the deviation between corresponding test result data is satisfied by predetermined threshold value respectively；
Step 111：If at least being deposited in the third physical base large-signal model and the 4th physical base large-signal model
In the simulation result of a physical base large-signal model the default threshold is still cannot be satisfied after only to physical parameter adjustment
Value, then readjust the target fitting parameter；
Step 112：If the emulation knot of the third physical base large-signal model and the 4th physical base large-signal model
Fruit is being satisfied by the predetermined threshold value after physical parameter adjustment, then is established with batch nitrogen using the target fitting parameter
The physical base model for changing gallium device, to which instruct device designs.
Fig. 5 is two arbitrarily chosen the nitrogen of gallium nitride device physical base large-signal model method for building up embodiment of the present invention
Change the simulation result and measured result comparison diagram of gallium device.
Wherein Fig. 5 (a) is the simulation result and measured result comparison diagram of one of gallium nitride device, and Fig. 5 (b) is another
The simulation result of a gallium nitride device and measured result comparison diagram.
Referring to Fig. 5, verification result of the invention shows simulation result and measured result more close to, it is seen that present invention gained
The reasonability of the fitting parameter arrived is higher.
Fig. 6 is the system construction drawing that gallium nitride device physical base large-signal model of the present invention establishes system embodiment.
Referring to Fig. 6, which establishes system, including：
Acquisition module 601, for obtaining to each microwave gallium nitride device to be measured in same processing line different defeated
Enter the test result data tested under power.
Mean value computation module 602, the mean value for calculating the test result data under each input power.
Fitting module 603 establishes initial physical base large-signal model for the mean value based on the test result data, and
The physical parameter that the initial physical base large-signal model is arranged is nominal value, obtains fitting parameter initial value.
Edge devices screening module 604, the test result data institute shape for screening each microwave gallium nitride device
At curve in the hithermost curve of envelope top edge curve and with the hithermost curve of envelope lower edge curve, obtain top
Edge device curves and lower edge device curves；The top edge device is the microwave corresponding to the top edge device curves
Gallium nitride device；The lower edge device is the microwave gallium nitride device corresponding to the lower edge device curves.
First model building module 605, for establishing the first physical base large-signal model based on the top edge device
With the second physical base large-signal model based on the lower edge device.
Model parameter determining module 606, for the fitting parameter initial value to be substituted into the big signal mode of the first physical base
Type and the second physical base large-signal model, and fitting parameter to the first physical base large-signal model and physical parameter and
The fitting parameter and physical parameter of the second physical base large-signal model are finely adjusted, and ensure the big letter of the first physical base
Number model and the second physical base large-signal model share same group of fitting parameter and physical parameter is different, make simulation result with
Deviation between the top edge device and the test result data of the lower edge device meets predetermined threshold value, to obtain mesh
Mark fitting parameter and target physical parameter.
The model parameter determining module 606, specifically includes：
Parameter substitutes into unit, for the fitting parameter initial value to be substituted into the first physical base large-signal model and described second
In physical base large-signal model；
Physical parameter adjustment unit, for adjusting the top edge device and the lower edge device in default adjusting range
The physical parameter of part judges the simulation result of the first physical base large-signal model and the second physical base large-signal model
Whether the deviation between corresponding test result data is satisfied by predetermined threshold value respectively, obtains judging result；
First parameter determination unit is tied if indicating that the deviation meets the predetermined threshold value for the judging result
Beam parameter adjustment determines that the physical parameter after the fitting parameter initial value and fine tuning is target fitting parameter and target physical
Parameter；
Second parameter determination unit, it is micro- if indicating that the deviation exceeds the predetermined threshold value for the judging result
The fitting parameter initial value is adjusted, the emulation of the first physical base large-signal model and the second physical base large-signal model is made
As a result the deviation respectively between corresponding test result data is satisfied by predetermined threshold value, and determines the fitting ginseng after fine tuning
Number and the physical parameter are target fitting parameter and target physical parameter.
Random screening module 607 is used for from the nitridation in addition to the top edge device and the lower edge device
Two gallium nitride devices are arbitrarily chosen in gallium device, obtain two gallium nitride devices to be verified.
Second model building module 608, the big signal mode of physical base for establishing two gallium nitride devices to be verified
Type obtains third physical base large-signal model and the 4th physical base large-signal model.
Substitute into module 609, for by the target fitting parameter substitute into respectively the third physical base large-signal model and
The 4th physical base large-signal model only adjusts the third physical base large-signal model and the big signal of the 4th physical base
The physical parameter of model obtains the emulation knot of the third physical base large-signal model and the 4th physical base large-signal model
Fruit.
Judgment module 610, for judging the third physical base large-signal model and the big signal mode of the 4th physical base
Whether the deviation between corresponding test result data is satisfied by predetermined threshold value to the simulation result of type respectively.
Feedback module 611, if being used for the third physical base large-signal model and the 4th physical base large-signal model
In at least exist a physical base large-signal model simulation result only to physical parameter adjustment after still cannot be satisfied institute
Predetermined threshold value is stated, then readjusts the target fitting parameter.
Application module 612, if being used for the third physical base large-signal model and the 4th physical base large-signal model
Simulation result by physical parameter adjustment after be satisfied by the predetermined threshold value, then using the target fitting parameter establish
With the physical base model of batch gallium nitride device, to which instruct device designs.
The method and system of disclosure of the invention have the following technical effects：
First, the mean value of test sample data is utilized when extracting fitting parameter initial value in the present invention, can be with fixed physical
Parameter extraction fitting parameter had both weakened the physical significance of fitting parameter in this way, at the beginning of more believable fitting parameter of having got back
Value.
Second, the present invention proposes the accurate extracting method of fitting parameter in physical base large-signal model, is joined according to physics
Several fluctuations extracts empirical parameter.The device for choosing test sample data envelope edge is modeled, and is fitted using same group
Parameter and fluctuation physical parameter, not only ensure that the empirical parameter of extraction was unrelated with device technology, but also can obtain accurate physics ginseng
Number.
Third, parameter extracting method of the invention can be used for establishing large batch of device model, because using same group
Fitting parameter, it is ensured that the physical parameter fluctuation of device model reflects true processing line fluctuation, and object is based on to establish
The big signal statistics model for managing basic mode type, to realize monitoring and optimization to processing line.
Principle and implementation of the present invention are described for specific case used herein, and above example is said
The bright method and its core concept for being merely used to help understand the present invention；Meanwhile for those of ordinary skill in the art, foundation
The thought of the present invention, there will be changes in the specific implementation manner and application range.In conclusion the content of the present specification is not
It is interpreted as limitation of the present invention.
Claims (6)
1. a kind of microwave gallium nitride device physical base large-signal model method for building up, which is characterized in that including：
Acquisition is tested to obtain to each microwave gallium nitride device to be measured in same processing line under different input power
Test result data；
Calculate the mean value of the test result data under each input power；
Mean value based on the test result data establishes initial physical base large-signal model, and it is big that the initial physical base is arranged
The physical parameter of signal model is nominal value, obtains fitting parameter initial value；
The test result data for screening each microwave gallium nitride device is formed by curve with envelope top edge curve most
Close curve and with the hithermost curve of envelope lower edge curve, obtain top edge device curves and lower edge device curves；
The top edge device is the microwave gallium nitride device corresponding to the top edge device curves；The lower edge device is
The microwave gallium nitride device corresponding to the lower edge device curves；
Establish the first physical base large-signal model based on the top edge device and the second object based on the lower edge device
Manage base large-signal model；
The fitting parameter initial value is substituted into the first physical base large-signal model and the second physical base large-signal model,
And fitting parameter to the first physical base large-signal model and physical parameter and the second physical base large-signal model
Fitting parameter and physical parameter be finely adjusted, and ensure that the first physical base large-signal model and second physical base are big
Signal model shares same group of fitting parameter and physical parameter is different, makes simulation result and the top edge device and described following
Deviation between the test result data of edge device meets predetermined threshold value, to obtain target fitting parameter and target physical ginseng
Number.
2. a kind of microwave gallium nitride device physical base large-signal model method for building up according to claim 1, feature exist
In described by fitting parameter initial value substitution the first physical base large-signal model and the big signal of the second physical base
Model, and fitting parameter to the first physical base large-signal model and physical parameter and the big signal of the second physical base
The fitting parameter and physical parameter of model are finely adjusted, and ensure the first physical base large-signal model and second physics
Base large-signal model shares same group of fitting parameter and physical parameter is different, makes simulation result and the top edge device and described
Deviation between the test result data of lower edge device meets predetermined threshold value, to obtain target fitting parameter and target physical
Parameter further includes：
Two nitridations are arbitrarily chosen from the gallium nitride device in addition to the top edge device and the lower edge device
Gallium device obtains two gallium nitride devices to be verified；
The physical base large-signal model for establishing two gallium nitride devices to be verified, obtain third physical base large-signal model and
4th physical base large-signal model；
The target fitting parameter is substituted into the third physical base large-signal model and the big signal of the 4th physical base respectively
Model only adjusts the physical parameter of the third physical base large-signal model and the 4th physical base large-signal model, obtains
The simulation result of the third physical base large-signal model and the 4th physical base large-signal model；
Judge the simulation result of the third physical base large-signal model and the 4th physical base large-signal model respectively with it is right
Whether the deviation between the test result data answered is satisfied by predetermined threshold value；
If at least there is a physical base in the third physical base large-signal model and the 4th physical base large-signal model
The simulation result of large-signal model still cannot be satisfied the predetermined threshold value after only to physical parameter adjustment, then readjusts
The target fitting parameter；
If the simulation result of the third physical base large-signal model and the 4th physical base large-signal model is by physics
It is satisfied by the predetermined threshold value after parameter adjustment, then establishes the object with batch gallium nitride device using the target fitting parameter
Basic mode type is managed, to the design of instruct device.
3. a kind of microwave gallium nitride device physical base large-signal model method for building up according to claim 1, feature exist
In, the fitting parameter initial value is substituted into the first physical base large-signal model and the second physical base large-signal model,
And fitting parameter to the first physical base large-signal model and physical parameter and the second physical base large-signal model
Fitting parameter and physical parameter be finely adjusted, and ensure that the first physical base large-signal model and second physical base are big
Signal model shares same group of fitting parameter and physical parameter is different, makes simulation result and the top edge device and described following
Deviation between the test result data of edge device meets predetermined threshold value, to obtain target fitting parameter and target physical ginseng
Number, specifically includes：
The fitting parameter initial value is substituted into the first physical base large-signal model and the second physical base large-signal model；
The physical parameter that the top edge device and the lower edge device are adjusted in default adjusting range, judges described first
The simulation result of physical base large-signal model and the second physical base large-signal model respectively with corresponding test result data
Between deviation whether be satisfied by predetermined threshold value, obtain judging result；
If the judging result indicates that the deviation meets the predetermined threshold value, terminate parameter adjustment, determines the fitting ginseng
The physical parameter after number initial value and fine tuning is target fitting parameter and target physical parameter；
If the judging result indicates that the deviation exceeds the predetermined threshold value, the fitting parameter initial value is finely tuned, is made described
The simulation result of first physical base large-signal model and the second physical base large-signal model respectively with corresponding test result
Deviation between data is satisfied by predetermined threshold value, and determines that the fitting parameter after fine tuning and the physical parameter are quasi- for target
Close parameter and target physical parameter.
4. a kind of microwave gallium nitride device physical base large-signal model establishes system, which is characterized in that including：
Acquisition module, for obtaining to each microwave gallium nitride device to be measured in same processing line under different input power
The test result data tested；
Mean value computation module, the mean value for calculating the test result data under each input power；
Fitting module establishes initial physical base large-signal model for the mean value based on the test result data, and institute is arranged
The physical parameter for stating initial physical base large-signal model is nominal value, obtains fitting parameter initial value；
Edge devices screening module, the test result data for screening each microwave gallium nitride device are formed by curve
In with the hithermost curve of envelope top edge curve and with the hithermost curve of envelope lower edge curve, it is bent to obtain top edge device
Line and lower edge device curves；The top edge device is the microwave gallium nitride device corresponding to the top edge device curves
Part；The lower edge device is the microwave gallium nitride device corresponding to the lower edge device curves；
First model building module, for establishing the first physical base large-signal model based on the top edge device and being based on institute
State the second physical base large-signal model of lower edge device；
Model parameter determining module, for the fitting parameter initial value to be substituted into the first physical base large-signal model and described
Second physical base large-signal model, and fitting parameter to the first physical base large-signal model and physical parameter and described
The fitting parameter and physical parameter of second physical base large-signal model are finely adjusted, and ensure the big signal mode of the first physical base
Type and the second physical base large-signal model share same group of fitting parameter and physical parameter is different, make simulation result with it is described
Deviation between top edge device and the test result data of the lower edge device meets predetermined threshold value, quasi- to obtain target
Close parameter and target physical parameter.
5. a kind of microwave gallium nitride device physical base large-signal model according to claim 4 establishes system, feature exists
In, the system also includes：
Random screening module, for from the gallium nitride device in addition to the top edge device and the lower edge device
Two gallium nitride devices are arbitrarily chosen, two gallium nitride devices to be verified are obtained；
Second model building module, the physical base large-signal model for establishing two gallium nitride devices to be verified, obtains
Third physical base large-signal model and the 4th physical base large-signal model；
Module is substituted into, for the target fitting parameter to be substituted into the third physical base large-signal model and the described 4th respectively
Physical base large-signal model only adjusts the object of the third physical base large-signal model and the 4th physical base large-signal model
Parameter is managed, the simulation result of the third physical base large-signal model and the 4th physical base large-signal model is obtained；
Judgment module, the emulation for judging the third physical base large-signal model and the 4th physical base large-signal model
As a result whether the deviation between corresponding test result data is satisfied by predetermined threshold value respectively；
Feedback module, if at least being deposited in the third physical base large-signal model and the 4th physical base large-signal model
In the simulation result of a physical base large-signal model the default threshold is still cannot be satisfied after only to physical parameter adjustment
Value, then readjust the target fitting parameter；
Application module, if the emulation knot for the third physical base large-signal model and the 4th physical base large-signal model
Fruit is being satisfied by the predetermined threshold value after physical parameter adjustment, then is established with batch nitrogen using the target fitting parameter
The physical base model for changing gallium device, to which instruct device designs.
6. a kind of microwave gallium nitride device physical base large-signal model according to claim 4 establishes system, feature exists
In the model parameter determining module specifically includes：
Parameter substitutes into unit, for the fitting parameter initial value to be substituted into the first physical base large-signal model and second physics
In base large-signal model；
Physical parameter adjustment unit, for adjusting the top edge device and the lower edge device in default adjusting range
Physical parameter judges the defeated simulation result point of the first physical base large-signal model and the second physical base large-signal model
Whether the deviation not between corresponding test result data is satisfied by predetermined threshold value, obtains judging result；
First parameter determination unit terminates to join if indicating that the deviation meets the predetermined threshold value for the judging result
Number adjustment determines that the physical parameter after the fitting parameter initial value and fine tuning is that target fitting parameter and target physical are joined
Number；
Second parameter determination unit finely tunes institute if indicating that the deviation exceeds the predetermined threshold value for the judging result
Fitting parameter initial value is stated, the simulation result of the first physical base large-signal model and the second physical base large-signal model is made
The deviation between corresponding test result data is satisfied by predetermined threshold value respectively, and determine the fitting parameter after fine tuning and
The physical parameter is target fitting parameter and target physical parameter.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810227481.8A CN108416155B (en) | 2018-03-20 | 2018-03-20 | Microwave gallium nitride device physical-based large signal model establishing method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810227481.8A CN108416155B (en) | 2018-03-20 | 2018-03-20 | Microwave gallium nitride device physical-based large signal model establishing method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108416155A true CN108416155A (en) | 2018-08-17 |
| CN108416155B CN108416155B (en) | 2020-03-24 |
Family
ID=63132933
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810227481.8A Active CN108416155B (en) | 2018-03-20 | 2018-03-20 | Microwave gallium nitride device physical-based large signal model establishing method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108416155B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109241623A (en) * | 2018-09-06 | 2019-01-18 | 电子科技大学 | A kind of surface potential compact model parameters extracting method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8044432B2 (en) * | 2005-11-29 | 2011-10-25 | The Hong Kong University Of Science And Technology | Low density drain HEMTs |
| EP2634809A2 (en) * | 2012-03-02 | 2013-09-04 | Kabushiki Kaisha Toshiba | Field effect transistor |
| CN104573330A (en) * | 2014-12-20 | 2015-04-29 | 电子科技大学 | I-V (Current-voltage) model parameter extraction method based on gallium nitride high electronic mobility crystal valve |
| CN105046066A (en) * | 2015-07-02 | 2015-11-11 | 中航（重庆）微电子有限公司 | AlGaN/GaN HETM small-signal model and parameter extraction method thereof |
| CN105825005A (en) * | 2016-03-15 | 2016-08-03 | 西安电子科技大学 | Construction method for nonlinear scalable GaN HEMT model |
| CN107636656A (en) * | 2016-07-12 | 2018-01-26 | 电子科技大学 | GaN device technological parameter statistical analysis technique based on large signal equivalent circuit model |
-
2018
- 2018-03-20 CN CN201810227481.8A patent/CN108416155B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8044432B2 (en) * | 2005-11-29 | 2011-10-25 | The Hong Kong University Of Science And Technology | Low density drain HEMTs |
| EP2634809A2 (en) * | 2012-03-02 | 2013-09-04 | Kabushiki Kaisha Toshiba | Field effect transistor |
| CN104573330A (en) * | 2014-12-20 | 2015-04-29 | 电子科技大学 | I-V (Current-voltage) model parameter extraction method based on gallium nitride high electronic mobility crystal valve |
| CN105046066A (en) * | 2015-07-02 | 2015-11-11 | 中航（重庆）微电子有限公司 | AlGaN/GaN HETM small-signal model and parameter extraction method thereof |
| CN105825005A (en) * | 2016-03-15 | 2016-08-03 | 西安电子科技大学 | Construction method for nonlinear scalable GaN HEMT model |
| CN107636656A (en) * | 2016-07-12 | 2018-01-26 | 电子科技大学 | GaN device technological parameter statistical analysis technique based on large signal equivalent circuit model |
Non-Patent Citations (2)
| Title |
|---|
| SHEIKH AAMIR AHSAN等: ""Physics-Based Multi-Bias RF Large-Signal GaN HEMT Modeling and Parameter Extraction Flow"", 《IEEE JOURNAL OF THE ELECTRON DEVICES SOCIETY》 * |
| 骆丹婷: "《氮化镓高电子迁移率晶体管的参数提取与灵敏度分析》", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109241623A (en) * | 2018-09-06 | 2019-01-18 | 电子科技大学 | A kind of surface potential compact model parameters extracting method |
| CN109241623B (en) * | 2018-09-06 | 2022-10-14 | 电子科技大学 | Surface potential compact model parameter extraction method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108416155B (en) | 2020-03-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9485040B2 (en) | Method for testing sensitivity of a data packet signal transceiver | |
| Lee et al. | A novel approach to extracting small-signal model parameters of silicon MOSFET's | |
| CN106372357B (en) | A kind of GaN HEMT nonlinear noise method for establishing model | |
| CN107636656A (en) | GaN device technological parameter statistical analysis technique based on large signal equivalent circuit model | |
| CN106845025B (en) | The big signal statistics model modelling approach of GaN high electron mobility transistor | |
| CN105760604B (en) | The modeling method of statistical model based on domain kindred effect | |
| CN100536320C (en) | Self-adaptive grid-voltage compensator and its compensation | |
| CN108416155A (en) | A kind of microwave gallium nitride device physical base large-signal model method for building up and system | |
| Wolff et al. | Heterogeneities in electricity grids strongly enhance non-Gaussian features of frequency fluctuations under stochastic power input | |
| Beleniotis et al. | An efficient drain-lag model for microwave GaN HEMTs based on ASM-HEMT | |
| CN106909741A (en) | A kind of modeling method of microwave GaN power devices | |
| CN112069759B (en) | GaN transistor modeling method and device based on Angelov model | |
| Bosi et al. | Gate waveform effects on high-efficiency PA design: An experimental validation | |
| Navarro et al. | Measurement-based ray-tracing models calibration in urban environments | |
| Kacmajor et al. | Microwave filter tuning for different center frequencies based on Artificial Neural Network and phase compensation | |
| KR20060108203A (en) | Method for testing the error ratio ber of a device according to confidence level, test time and selectivity | |
| Green et al. | Validation of an analytical large signal model for AlGaN/GaN HEMTs | |
| US11703537B2 (en) | Method and apparatus for analysis of interface state of MIS-HEMT device | |
| Patnaik et al. | Implementation of a 18 GHz bandwidth Channel Emulator using an FIR filter | |
| CN106254001A (en) | A kind of SRS signal channel quality measuring method and device | |
| CN104750899B (en) | The parameter extracting method of bipolar transistor | |
| CN104134862B (en) | Circular patch tie Microstrip Antenna computational methods | |
| WO2022088042A1 (en) | Method and apparatus for measuring nonlinear relevant parameter of nonlinear device, and system | |
| US11153844B2 (en) | Method for estimating position of wireless network equipment | |
| El Furqany et al. | Spectral analysis and SARIMA model for forecasting Indian Ocean Dipole (IOD) and rainfall in West Aceh Regency |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |